View source for 17495394. METHOD FOR ENRICHED TRAINING BY USING PREDICTED FEATURES OBTAINED FROM MULTIPLE MODELS simplified abstract (International Business Machines Corporation)

=METHOD FOR ENRICHED TRAINING BY USING PREDICTED FEATURES OBTAINED FROM MULTIPLE MODELS=


==Organization Name==

[[:Category:International Business Machines Corporation|International Business Machines Corporation]]


[[Category:International Business Machines Corporation]]

==Inventor(s)==

[[:Category:Simona Rabinovici-cohen of Haifa (IL)|Simona Rabinovici-cohen of Haifa (IL)]][[Category:Simona Rabinovici-cohen of Haifa (IL)]]

[[:Category:SHAKED Perek of Givatayim (IL)|SHAKED Perek of Givatayim (IL)]][[Category:SHAKED Perek of Givatayim (IL)]]

[[:Category:Tal Tlusty Shapiro of Zichron Yaacov (IL)|Tal Tlusty Shapiro of Zichron Yaacov (IL)]][[Category:Tal Tlusty Shapiro of Zichron Yaacov (IL)]]

[[:Category:EFRAT Hexter of Beit Shemesh (IL)|EFRAT Hexter of Beit Shemesh (IL)]][[Category:EFRAT Hexter of Beit Shemesh (IL)]]

==METHOD FOR ENRICHED TRAINING BY USING PREDICTED FEATURES OBTAINED FROM MULTIPLE MODELS - A simplified explanation of the abstract==


This abstract first appeared for US patent application 17495394 titled 'METHOD FOR ENRICHED TRAINING BY USING PREDICTED FEATURES OBTAINED FROM MULTIPLE MODELS

==Simplified Explanation==

The patent application describes a system and method for training machine learning models by using predicted features from other models, third party models, and legacy models to augment training datasets.

* The method involves using additional machine learning models to predict unknown features in datasets.
* The predicted features are used to generate a distribution of scores and create replicas of items with missing features.
* The replicas have different estimations of the unknown features.
* The machine learning model is then trained using the known features and the scores for the unknown feature of each item.

Potential applications of this technology:

* Improving the accuracy and reliability of machine learning models by augmenting training datasets with predicted features.
* Enhancing the performance of models in various domains such as image recognition, natural language processing, and recommendation systems.

Problems solved by this technology:

* Many datasets have missing or erroneous values for features, which can negatively impact the performance of machine learning models.
* Predicting unknown features and generating replicas with different estimations can help address the issue of missing data and improve the training process.

Benefits of this technology:

* Enables the use of predicted features from other models to enhance the training of machine learning models.
* Increases the robustness and accuracy of models by incorporating additional information from various sources.
* Provides a method for handling missing data in datasets and improving the overall performance of machine learning models.


==Original Abstract Submitted==

A system and a method for training machine learning models, using features predicted by other models, third party models, legacy models, and the like, for training dataset augmentation. Many datasets have many items with unknown, missing, or erroneous values of features. The method comprises using additional machine learning models to predict unknown features, optionally generate a distribution from their inferred score, and use a plurality of scores from the distribution and/or scores from further additional machine learning models, to create replicas for item with missing features, having different estimations of the unknown features. Followingly, train the machine learning model using data items with the known features and the scores for the unknown feature of the item.



[[Category:G06K9/62]]
[[Category:G06N20/20]]